> For the complete documentation index, see [llms.txt](https://context-data.gitbook.io/context-data-1/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://context-data.gitbook.io/context-data-1/quickstart.md).

# Quickstart

<figure><img src="/files/JbPpmym2xcqWdcrm4IYR" alt=""><figcaption></figcaption></figure>

**There are four basic steps to build an end-to-end flow on Context Data**

1\). Source Connection: Build connection(s) to where your source data resides (e.g. MySQL, PostgreSQL, Amazon S3)

2\). Embedding Model: Create a link to the embedding model which will convert data retrieved from the source to vector embeddings (basically an array of numbers)

3\). Target Connection: Build connection(s) to where the vector embeddings will be saved (and where your AI application will read from)

4\). Flow: The flow ties of the steps above (source connection, embedding model and target connection) into an end-to-end process ready to be executed.

Basically, when a flow is triggered, it will:

* Get the data from the source connection that you defined
* Convert the retrieved data to a format optimized for vector search
* Write the converted data to the vector database/store&#x20;

{% embed url="<https://www.youtube.com/watch?v=7wXk_kATjFQ>" %}